Genderedness and Valence

survey results!

I asked people to rate 266 items on a scale from how masculine to feminine they were. I included lots of stuff - personality traits, abstract concepts, moods, body parts, random items - stuff that was both clearly gendered and ambiguous.

I then asked people to rate the same set of items on a scale from goodness to badness.

Here’s the total results, plotted, normalized by sex.

(Raw averages are here, in case you want to make better plots yourself, and the complete raw data (with some demographic info removed) is here)

An average of 800 people answered each question for the masculinity/femininity survey, and 280 answered each good/bad (valence) question.

Full size image with zoom is here

There’s still a lot, so it’s hard to read. So I’m gonna take some subsections!

Here’s just personality or personhood type traits, actions, vibes:

Here’s stuff related to sex, gender, or body parts:

Types of people:

Places ish

Concrete themes

Politics

Abstract things

I didn’t spend a lot of time making these charts super chic; I might go through and manually fancy up the very first/big one. My apologies if you notice any artifacts! I also wasn’t super precise with what items I included in what chart.

Also the red-blue isn’t the most intuitive because it’s also gendered color, I know, I’m sorry, I just don’t want to redo all the charts again.

If it wasn’t clear - the color spectrum is an indication of how men and women taking the survey, rated things differently.

So for example, women rated ‘wisdom’ as quite feminine, while men rated it as quite masculine. This means people disproportionately assigned it to their own gender.

Whereas women rated ‘blowjobs’ as quite masculine, and men rated ‘blowjobs’ as quite feminine. This means people disproportionately assigned it to the opposite gender.

The correlation between the average rated valence (how good/bad something was) and how much men and women disagreed about it, assigning it to their own gender, was r=0.3.

In general, there was a slight correlation between how masculine an item was, and how ‘bad’ people rated it. Not sure if this is due to me subconsciously adding more bad-masculine things and more good-feminine things though!

---

Men and women also disagreed on their ratings on how good/bad things were.

The greatest disagreements where women rated items as higher than men did:

1. astrology (gap of 23)

2. crystals meant for healing (19)

3. tarot cards (15)

4. sexual interest in men (15)

5. The song Toxic, by Britney Spears (13)

6. a snake (13)

7. chandeliers (12)

8. socialism (12)

9. spirituality (12)

10. A bear (12)

Whereas these are the ones men rated higher than women:

1. math (gap of 13)

2. capitalism (13)

3. casual sex (12)

4. pregnancy (11)

5. blowjobs (11)

6. your mother (11)

7. youth (9)

8. civilization (9)

9. a very rich person (7)

10. Mario (the video game character (7)

---

Here’s the top differences in gender ratings.

These are ones that the genders fingerpointed at the other gender the most - women rated as most masculine, while men rated as most feminine:

1. entitlement (gap of 17)

2. blowjobs (15)

3. romantic love (13)

4. sexual interest in men (12)

5. a trans man (11)

6. dildos (11)

7. cowardice (11)

8. narcissistic personality disorder (9)

9. North Korea (8)

10. incompetence (8)

Here’s the ones that men and women each claimed for themselves - women rated as most feminine, while men rated as most masculine.

1. your own genitals (gap of 67)

2. you yourself (gap of 42)

3. wisdom (19)

4. family love (17)

5. loyalty (14)

6. being reasonable (13)

7. friendship love (13)

8. curiosity (12)

9. emotional maturity (12)

10. perceptiveness (12)

The survey design was basically that I asked people some demographic info, and then asked them to rate the items on a 0-100 scale from full femininity, to full masculine - and then separately, later, made a survey that was identical, except asked people to rate things on a “good/positive/nice” or “bad/negative/unpleasant” spectrum. I included 266 items, mostly randomly selected, and attempting to cover a broad range of things, but focusing on things that might be even slightly gender-related. People got the items at random, and were encouraged to quit whenever they got bored. I posted the survey on my twitter (so a bit more selection bias than most of my surveys!). Around ~2500 people took the gender survey, and 770 took the valence survey, 80% of them men.

In general, I think this kind of survey is slightly more robust than normal for lower samples; there’s not a ton of outliers, and I’m not measuring anything I need to know with high confidence rates of fringe populations, and I’m asking about a topic that tends to be kinda diffuse/reflected through consensus social reality. Still, though, I’d expect this to be really “western, youth” lens.

I added a few questions - fluffers, children, empathizing with others, social power, anxiety, luck, friend love, family love, romantic love, unconditional love - in last minute, and so they have a lower sample.